Abstract
According to process reliabilism, a belief produced by a reliable belief-forming process is justified. I introduce problems for this theory on any account of reliability. Does the performance of a process in some domain of worlds settle its reliability? The theories that answer “Yes” typically fail to state the temporal parameters of this performance. I argue that any theory paired with any plausible parameters has implausible implications. The theories that answer “No,” I argue, thereby lack essential support and exacerbate familiar problems. There are new reasons to avoid any reliability conditions on justification.
Similar content being viewed by others
Notes
Cf. Goldman (2008a).
AWR descends from a theory of the truth-conditions for statements about justification (cf. Comesaña 2002; Goldman 1992). For discussion see Becker (2007), Goldman (1986), and Lyons (2013). Goldman (2009) strongly favors AWR. And Goldman (2008a) seems to defend AWR–he discusses several versions of reliabilism, and leaves all but AWR with unanswered objections. When asked about AWR, Goldman said in conversation that he changes his mind about which reliabilist view is right.
Throughout, for ease of exposition I simplify reliabilism, omitting details about belief-dependent belief-forming processes, belief-sustaining processes, and prima facie justification. Nothing hangs on this.
Cf. Baumann (2009: 85–86). Why “most”? Because some processes only ever produce true beliefs, and others only ever produce false beliefs. Whether these processes are reliable is independent of temporal parameters.
Goldman (1986: 49) is ambiguous between this version of Opportune Reliability and Historical Reliability (see below): “We may take [reliability] to refer to the truth ratio of beliefs generated by the process on all actual occasions of use”. Does he mean all actual occasions there will ever be, or all actual occasions to date?
It is difficult to identify times in non-actual worlds with times in the actual world. Yet AWR must do this given any dynamic account of reliability. In the interests of charity, however, I will not assume such an identification is impossible. I thank an anonymous reviewer for noting the difficulty.
See Comesaña (2002: 259–260).
I thank several anonymous referees at other journals for this suggestion.
Notably, reliabilists like Henderson and Horgan (2010: 91) think that to identify the relevant processes in this way is unscientific and “really is a convenient proxy for what ultimately presumably matters”.
BonJour (2001/2010: 218–219) thinks externalist theories of justification can’t accommodate this because we lack direct access to the external factors that determine the justificatory status of beliefs like B or G. Kornblith (2004: 198) thinks externalism can accommodate. Yet BonJour and Kornblith agree that a theory of justification should accommodate our intuitions about our object-level justification. They are right. Epistemology seems hopeless otherwise. Epistemological theories seem meritless if they are selected to fit meritless intuitive judgments about justification. However, unlike BonJour, I am pointing out what follows if AWR can accommodate the justification of the judgments.
One can evaluate reliabilism only by stipulating which process of belief-formation is relevant. For my purposes it ultimately does not matter which ones we stipulate. Comesaña (2002) and Sosa (2003) take taking perception at face value to be the relevant process that forms beliefs like B (perhaps this can be spelled out as forming beliefs that fit the representational contents of perceptual experience). Cf. Goldman (2008a, 1986, 1979), who moreover takes guessing to be the relevant process that forms beliefs like G. An adequate solution to the generality problem may individuate relevant processes at a finer grain than these.
Blindluck resembles an example that presents what I’ll call the clairvoyance problem for reliabilism: some possible subject without any relevant information forms beliefs by a reliable process, and reliabilism implausibly counts these beliefs as justified (see BonJour 1980). Experience Machine resembles an example that presents the new evil demon problem for reliabilism: a massively deceived possible subject who is mentally just like me should also share all my justification, but reliabilism implausibly implies otherwise since just one of us reliably forms beliefs (see Cohen 1984). However, Blindluck and Experience Machine present new problems because they, unlike the examples, concern actual beliefs and concern the reliability of processes in the actual world, processes whose alleged actual tendencies have been thought to motivate reliabilism.
Notably, most epistemologists of testimony agree with Lackey (2006: 440) that “the observational base of ordinary epistemic agents is simply far too small to allow the requisite induction about the reliability of testimony.” But our observational base for testimony is not relevantly smaller than our bases for perception, guessing, memory, hasty generalization, etc. If Lackey is right, then the consequents in conclusions like (C) and (C*) lack support, and so AWR lacks support. If Lackey is wrong, then the current orthodoxy about testimony includes an importantly false assumption.
A variant of Cohen’s (2002) problem of easy knowledge.
Levin (1997) endorses Same World Reliabilism. For discussion see Lyons (2013). Sosa (2001) defends a view that entails both AWR and Same World Reliabilism but that assigns them distinct concepts of justification. Comesaña (2010, 2002) defends Indexical Reliabilism, a view intended to express both AWR and Same World Reliabilism; Ball and Blome-Tillman (2013) criticize it. On Indexical Reliabilism, the truth-value of all propositions expressed by a justification attribution depends on either of two worlds (the actual world and the subject’s world), so it is not clearly a version of Lone World Reliabilism. At any rate, Indexical Reliabilism faces the main problems AWR and Same World Reliabilism face here since, on it, a process’ actual performance determines both its reliability and its justificatory powers in the actual world (see below). Becker (2007: 32) attacks what he calls “actual world reliabilism” by citing an example of Plantinga’s (1993: 199). Greco (2000: 175–176) also uses the example, but neither he nor Plantinga specifies whether it is meant to threaten Same World Reliabilism or AWR. The example involves a non-actual world, such that it threatens only Same World Reliabilism.
Henderson and Horgan (2010: 114–115) propose other conditions, centered on “degrees of variation” across the worlds in the domain, such that their version of Modal Reliabilism faces Probabilistic Reliabilism’s reliability problem (see below).
Comesaña (2006) attempts to solve the generality problem by identifying the relevant process as basing belief that p on e. But given his solution, contingent external factors cannot even affect which process is relevant. On his solution, Logical Reliabilism could remain externalist if these factors nonetheless determine which of a subject’s mental features count as evidence for p. But if evidential support relations are necessary, then Comesaña’s Logical Reliabilism paired with his solution to the generality problem would deny externalism. For further criticism of Comesaña’s solution, see Matheson (2015).
The reader may have noticed that Comesaña (2010) endorses not only Logical Reliabilism but also Indexical Reliabilism, which expresses both Same World Reliabilism and AWR: roughly, S’s belief that p is justified iff the type producing a belief that p based on e is actually reliable. Depending on how ‘actually’ is understood, this expresses either of two propositions—the horizontal or the diagonal proposition in 2-dimensional semantics. Comesaña says that the truth-values of the diagonal and horizontal propositions may differ. This is supposed to explain our allegedly conflicting intuitions about the deceived victim in the new evil demon problem; the victim’s beliefs are justified (as AWR implies) because they are formed by processes that are reliable in the actual world, but they are unjustified (as Same World Reliabilism implies) because they are formed by processes that are not reliable in the victim’s world.
Comesaña’s reliabilism-cocktail has consequences he might not like. On Logical Reliabilism, the relevant probability relations are necessary—if a process is reliable in some world, it’s reliable in all. So, the diagonal and horizontal propositions expressed by Indexical Reliabilism are logically equivalent. AWR and Same World Reliabilism, then, are identical. So the cocktail can’t explain conflicting intuitions about the deceived victim.
References
Alston, W. (1988). An internalist externalism. Synthese, 74, 265–283.
Alston, W. (1995). How to think about reliability. Philosophical Topics, 23(1), 1–29.
Ball, B., & Blome-Tillman, M. (2013). Indexical reliabilism and the new evil Demon. Erkenntnis, 78(6), 1317–1336.
Baumann, P. (1998). Can reliabilists believe in subjective probability? The Philosophical Quarterly, 48, 199–200.
Baumann, P. (2009). Reliabilism—Modal, probabilistic, or contextualist. Grazer Philsophische Studien, 79, 77–89.
Becker, K. (2007). Epistemology modalized. New York: Routledge.
Bergmann, M. (2006). Justification without awareness. Oxford: Oxford University Press.
BonJour, L. (1980). Externalist theories of empirical knowledge. Midwest Studies in Philosophy, 5, 53–73.
BonJour, L. (2001/2010). Epistemology: Classic problems and contemporary responses. Lanham: Rowman & Littlefield.
Cohen, S. (1984). Justification and truth. Philosophical Studies, 46, 279–295.
Cohen, S. (2002). Bootstrapping and the problem of easy knowledge. Philosophy and Phenomenological Research, 65(2), 309–329.
Comesaña, J. (2002). The diagonal and the Demon. Philosophical Studies, 110, 246–266.
Comesaña, J. (2006). A well-founded solution to the generality problem. Philosophical Studies, 129(1), 27–47.
Comesaña, J. (2010). Evidentialist reliabilism. Noûs, 44(4), 571–600.
Conee, E., & Feldman, R. (1998). The generality problem for reliabilism. Philosophical Studies, 89, 1–29.
Feldman, R. (1985). Reliability and justification. The Monist, 64(1), 59–74.
Feldman, R., & Conee, E. (1985). Evidentialism. Philosophical Studies, 48, 15–34.
Feldman, R., & Conee, E. (2002). Typing problems. Philosophy and Phenomenological Research, 65(1), 98–105.
Frise, M. (2014). Speaking freely: On free will and the epistemology of testimony.Synthese, 191(7), 1587–1603.
Goldman, A. (1979). What is justified belief? In G. Pappas (Ed.), Justification and knowledge. Dordrecht: Reidel.
Goldman, A. (1986). Epistemology and cognition. Cambridge: Harvard University Press.
Goldman, A. (1988). Strong and weak justification. Philosophical perspectives, 2, 51–69.
Goldman, A. (1992). Epistemic folkway and scientific methodology. In A. Goldman (Ed.), Liasons: Philosophy meets the cognitive and social sciences (pp. 155–175). Cambridge: MIT Press.
Goldman, A. (2008a). Reliabilism. In E. Zalta (Ed.), Stanford encyclopedia of philosophy (Fall 2008 Edition). http://plato.stanford.edu/archives/fall2008/entrices/reliabilism/.
Goldman, A. (2008b). Immediate justification and process reliabilism. In Q. Smith (Ed.), Epistemology: New essays. Oxford: Oxford University Press.
Goldman, A. (2009). Replies to discussants. Grazer Philsophische Studien, 79, 245–288.
Greco, J. (1999). Agent reliabilism. Philosophical perspectives, 13, 273–296.
Greco, J. (2000). Putting skeptics in their place (p. 2000). Cambridge: Cambridge University Press.
Henderson, D., & Horgan, T. (2010). The epistemological spectrum. Oxford: Oxford University Press.
Hill, C. (1996). Process reliabilism and cartesian scepticism. Philosophy and Phenomenological Research, 56(3), 567–581.
Kornblith, H. (2004). Does reliabilism make knowledge merely conditional? Philosophical Issues, 14, 185–200.
Kvart, I. (2006). A probabilist theory of knowledge. Philosophy and Phenomenological Research, 77(1), 1–43.
Lackey, J. (2006). Knowing from testimony. Philosophy Compass, 1(5), 432–448.
Levin, M. (1997). You can always count on reliabilism. Philosophy and Phenomenological Research, 57(3), 607–617.
Lyons, J. (2013). Should reliabilists be worried about Demon worlds? Philosophy and Phenomenological Research, 86(1), 1–40.
Matheson, J. (2015). Is there a well-founded solution to the generality problem? Philosophical Studies, 172, 459–468.
Plantinga, A. (1993). Warrant: The current debate. Oxford: Oxford University Press.
Pollock, J. (1984). Reliability and justified belief. Canadian Journal of Philosophy, 14(1), 103–114.
Sosa, E. (2001). Goldman’s reliabilism and virtue epistemology. Philosophical Topics, 29(1–2), 383–400.
Sosa, E. (2003). Ontology, understanding, and the a priori. Ratio, 2, 178–188.
Vogel, J. (2000). Reliabilism leveled. Journal of Philosophy, 97(11), 602–623.
Acknowledgements
I am grateful to Jon Matheson, Alyssa Ney, Brad Rettler, Lindsay Rettler, Brett Sherman, three anonymous referees, and an audience at UT Austin for valuable discussion of this paper. Thanks especially to Earl Conee, Rich Feldman, and Kevin McCain for helpful comments on multiple drafts. I revised this paper while supported by a grant from the Templeton Religion Trust. The opinions expressed in this paper are my own and do not necessarily reflect the views of the Templeton Religion Trust.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Frise, M. The reliability problem for reliabilism. Philos Stud 175, 923–945 (2018). https://doi.org/10.1007/s11098-017-0899-0
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11098-017-0899-0